Add Y-label and Vertical Bar Label to Graph Options

[Maxence Morlet]

Hi all,

I am using rdpow from ssc. I am interested in modifying the graph that this user-written command produces through the graph_options option. Namely, I would like to label the values of the y-axis and add a vertical line to the x-axis. This vertical line should carry the label, visible on the graph itself, "Mean of Dep. Var." The vertical line should be at x=2.02.

Here is the code I've tried:

rdpow y running_var , tau(5) plot graph_options(xla(-7.5 -5 -2.5 0 2.5 5 7.5) yla(.1 "10%" .2 "20%" .3 "30%" .4 "40%" .5 "50%" .6 "60%") xaxis(1 2) xline(2.02) xlabel(2.02 "Mean of Dep. Var.", axis(2)) xtitle("",axis(2)) title(Power Function) xtitle(Magnitude of Effect to Detect) ytitle(Power)) covs($covs) vce(cluster county)

This is the error I got:

xlabels(2.02 Mean of Dep. Var. , axis(2)) not allowed, xaxis2 does not exist

When I remove

xaxis(1 2) xline(2.02) xlabel(2.02 "Mean of Dep. Var.", axis(2)) xtitle("",axis(2))

and only try to label the y axis, I get:

invalid label specifier, :  .1 10% .2 20% .3 30% .4 40% .5 50% .6 60%:

Could someone please let me know what I am doing wrong?

Thanks in advance!

[Jared Greathouse]

Have you gotten this idea to work on a simple data example in a do file? I'm no stranger to modifying others ado files, but before I do I ensure that whatever modification I desire works outside of the ado file first on a simple example, in a do file. As it stands though, I'm unsure at to why you get this error

[Nick Cox]

rdpow is from Stata Journal, not SSC.

Clearly the "" you're trying to pass. get stripped,

I would try cloning the program under a different name and replacing

graph_options(string)

in the syntax statement with

graph_options(string asis)

and any check for

"`graph_options'" == ""

with

`"`graph_options'"' == ""

[Maxence Morlet]

Dear both,

Thank you very much for your replies.

Jared Greathouse - You're right, the thing is for the results to make sense with my data, I would have had to post hundreds of rows of data. But now that you mention it, I should have used the rdrobust_senate file as dataset, from the seminal RDD study.

Nick Cox - I will give it a try and get back to you. Thank you for the suggestion!

[Andrew Musau]

Another option is to save the generated graph (GPH file) and edit it with addplot from SSC. See

ssc desc addplot